Changes

Summary

1. WIP PVG rewrite (commit: dc6464b) (details)
2. more massive refactoring (commit: 1b79e0c) (details)
3. checkpoint (commit: 08b233a) (details)
4. push an update (commit: 329d8fe) (details)
5. another checkpoint (commit: 8a375f9) (details)
6. checkpoint again. (commit: a37876e) (details)
7. WIP (commit: c728bac) (details)
8. mostly working (commit: d4d90db) (details)
9. more WIP (commit: 342d5a0) (details)
10. many fixes (commit: 0e68498) (details)
11. Code cleanup (commit: 51c6c15) (details)
12. fixed burn time fixes and test fixes (commit: 3dfe8a1) (details)
13. Fix optimizer trying to coast when it shouldn't (commit: 46d5fd7) (details)
14. Some code cleanup in the PVG Ascent algorithm (commit: 3c4b656) (details)
15. unrelated code cleanup (commit: 57b0624) (details)
16. Fix FuelFlowSim for NO_FLOW resources in Engines (commit: 46a9693) (details)
17. Fix StagingController for NO_FLOW fuels (commit: 42435f4) (details)
18. remove dead code comment (commit: 93f2709) (details)
19. simplify MJlib logger (commit: 9a0d17f) (details)
20. Use new ToSI extnesion method everywhere (commit: ba84b68) (details)
21. Try to see if JetBrains.Annotations breaks builds (commit: 97b02b9) (details)
22. Apply UsedImplicitly to fix dead code analysis (commit: e62cd5e) (details)
23. code cleanup to MechJebCore (commit: be96049) (details)
24. Use aggressive inlining (commit: df5aebc) (details)
25. Code cleanup for the Maneuver Operations (commit: 12ae2c6) (details)
26. Move ODE in MathJ to MechJebLib (commit: 2552971) (details)
27. Change License on ODE Solver (commit: 447f766) (details)
28. Move "SpaceMath" function into MechJebLib (commit: f5a5d40) (details)
29. M3 fixes and a lot more (commit: ff93427) (details)
30. Change "attach altitude" to "burnout altitude" (commit: 47daf5b) (details)
31. Change "stop at stage" to make it clearer it affects autostaging (commit: ad2c710) (details)
32. remove spammy logging (commit: eb80ae9) (details)
33. Address warnings (commit: d28bbe9) (details)
34. Fix PVG terminal staging bug (commit: adf3ecd) (details)
35. Fix the autostage limit (commit: 574c150) (details)
36. Track down and fix some annoying allocations in the ODE+interpolants (commit: 9763ffd) (details)
37. Add proper tests for Gooding Lambert calculator (commit: 2be3676) (details)
38. Remove SwapYZ API (commit: 81232eb) (details)
39. Automated cleanup of many ancillary source code modules (commit: 78d2fa5) (details)
40. Rename the "Swapped" API (commit: 6086f8e) (details)
41. clean up the DeltaVToInterceptAtTime() API. (commit: bb548ca) (details)
42. fix probable transfer calculator bug (commit: 2061c32) (details)
43. ReturnFromMoon reimplementation and large reorganization (commit: 4bb4091) (details)
44. Cleanup of Angle APIs (commit: 47adf58) (details)
45. Make Moon Return work for initial hyperbolic (commit: d5d7976) (details)
46. Use the perifocal frame for Apsis change. (commit: 7a8d75e) (details)
47. PID tuning (commit: ec33cd6) (details)
48. More PID tweaks (commit: 850d2d4) (details)
49. Some VesselState cleanup (commit: c7cbe95) (details)
50. Revert "Some VesselState cleanup" (commit: 0daaf4e) (details)
51. clean up the scale struct (commit: c900649) (details)
52. Reuse the scale object in the return to moon maneuver (commit: cde27ec) (details)
53. reorganize and extract the maneuvers in mechjeblib a little better (commit: 213512f) (details)
54. Redo SMA Changing Maneuver as SQP problem (commit: b3049f5) (details)
55. Reformat and Cleanup on a lot of code (commit: ffb635d) (details)
56. Add a maneuver to change eccentricity (commit: 77644d2) (details)
57. Drop the vel deadband (commit: 9dafa07) (details)
58. Fix some namespacing issues (commit: bc58481) (details)
59. Missed one namespace issue (commit: 9fa4f38) (details)
60. Maneuver tweaks (commit: e588c0b) (details)
61. Compound assingment in Gooding solver (commit: 2547f3e) (details)
62. DormandPrince overhaul (commit: a840017) (details)
63. better working error controller (commit: 3010d00) (details)
64. PI error control (commit: c73a9cc) (details)
65. Simplify how I was thinking about K states (commit: 9359ed2) (details)
66. Add exception handling to cleanup (commit: 0c6899f) (details)
67. Basic minstep/maxstep logic (commit: dd95067) (details)
68. rearrange the DP5 interpolant (commit: 2cb9292) (details)
69. Hang state off the IVP object (commit: d277897) (details)
70. Optimize FuelFlowSimulation (commit: 0402a17) (details)
71. Fix .vs folder not being correctly gitignored (commit: fc51317) (details)
72. Optimize Stage stats simulation by copying second FuelFlowSimulation from the first instead of initilizing them twice (commit: a82bc81) (details)
73. Make #1515 account for changes made in 1660 (commit: 944f9b2) (details)
74. Fix 2 cases where logging was done directly from background thread (commit: cd87ef9) (details)
75. Lot of work (commit: 51e44f9) (details)
76. Make terminal events work correctly, add a test (commit: d89220e) (details)
77. Fix PVG without delta-V window open (commit: 8775cca) (details)
78. ReentrySimulation code cleanup (commit: e59de6f) (details)
79. Extract SimCurves out to its own file (commit: f0e960a) (details)
80. Cleanup objectpool global/local case (commit: d57bc20) (details)
81. Convert to shape of Smooth.Pools API (commit: 1d7ebb1) (details)
82. Fix ToSI issue (commit: 6dc55c3) (details)
83. Add new metric prefixes (commit: a380a1f) (details)
84. ToSI tweaks (commit: 02fab0d) (details)
85. Fix PVG After stage (commit: ae6cea6) (details)
86. Analytic coast integrations with PVG (commit: 4be1e06) (details)
87. Update MechJebLib copyright headers (commit: 34a9667) (details)
88. Missed this in the License PR (commit: c6ca586) (details)
89. Fix sign in coast time update (commit: a208bbf) (details)
90. Terminal Guidance correctness bug (commit: 4cde5c3) (details)
91. Fix CoastBefore behavior (commit: 87984d6) (details)
92. Update alglib to 4.0 (commit: a32c207) (details)
93. PVG: Use Analytic for bootstrap and Integrator for converged (commit: 1547d2b) (details)
94. better error message for converged failure (commit: 1f961fe) (details)
95. Convert the ArrayWrapper class to a struct (commit: 26de45d) (details)
96. Fix autostaging for asparagus boosters (commit: d9e18de) (details)
97. Misc cleanup (commit: 9dfc8a8) (details)
98. Fix Fairing Separation logic (commit: 4392aaa) (details)
99. More PID tuning (commit: 0c1ff26) (details)
100. Fix caching bug in FuelFlowSimulation (commit: de3afff) (details)
101. Automated Code Cleanup (commit: 747cacf) (details)
102. Refactor ChangeOrbitalElements slightly (commit: 5cd9811) (details)
103. PVG cleanup: replace ResidualWrapper with ContinuityLayout/ResidualLayout (commit: 118a797) (details)
104. More PID tuning (commit: 2dbfcc6) (details)
105. Rename to "PVG Settings" and use horizontal split (commit: b0ee3b5) (details)
106. Stop fairing sep in a coast from ending the coast (commit: e6eabea) (details)
107. Splitup burnout altitude setting (commit: 333e5d5) (details)
108. Fix both checks that use IndexForKSPStage (commit: f3eac60) (details)
109. Optimize Stage is now Early Shutoff Stage (commit: 66836f5) (details)
110. Increase Pos deadband to 0.002 (commit: 1f51064) (details)
111. Remove the ScriptsModule (commit: 681d58f) (details)
112. Landing autopilot style cleanup (commit: 21844b9) (details)
113. Remove the MechJebPod part (commit: 00ac5bd) (details)
114. Update gitignore for JetBrains IDEs (commit: 4652221) (details)
115. Possibly breaking cleanup of MechJebCore (commit: 1bcbef0) (details)
116. ComputerModule breaking API changes (commit: 00891f8) (details)
117. Mostly automated FFS code cleanup (commit: da6dfe0) (details)
118. Add back the Ascent AP Toggle (commit: 6048d7c) (details)
119. ergh, typo (commit: cd5311f) (details)
120. Update to C# 9.0, add partial suicide burn overhaul (commit: c206ba6) (details)
121. Guess we can't do this bit of C#9 (commit: 8afb1ee) (details)
122. And sadly we can't use new() either (commit: abb138a) (details)
123. Back to C# 8.0 (commit: 863fb30) (details)
124. And fix the code to compile properly on C# 8.0 (commit: 9adc415) (details)
125. Savegame:  working simulation with residuals (commit: 55727c5) (details)
126. Savegame: algorithm is basically working (commit: 7f45d8b) (details)
127. support activating RCS (commit: d32fc91) (details)
128. Clean up API a bit (commit: f1d5eb3) (details)
129. Break up Build and Update and use object pooling (commit: 13fb2ac) (details)
130. Use callbacks and fix a few bugs (commit: 34eb79e) (details)
131. Savegame: lots of fixes (commit: c560944) (details)
132. Savegame: getting close (commit: 663f947) (details)
133. Finished with perf pass and more bugs fixed (commit: 8e3791d) (details)
134. slighly more readable reflection (commit: fcb566a) (details)
135. even more betterer reflection (commit: b381fcd) (details)
136. okay now we're getting concise (commit: 2b5ee16) (details)
137. Finished threading and everything done on the TODO list (commit: ec3be1b) (details)
138. Faster flight refresh interval (commit: ee36930) (details)
139. Cascade disabling of Guidance Controller to PVG Autopilot (commit: d9df5df) (details)
140. Let multiple modules limit autostaging and add/remove independently (commit: 7bb88a3) (details)
141. ThrustController and StagingController cleanup (commit: 29b2e36) (details)
142. add back adjusting apo-at-apo and per-at-per (commit: 9605873) (details)
143. Reset SpoolupCurrent to 0 on recalc (commit: cc8262c) (details)
144. make it pretty (commit: ca8cc5b) (details)
145. Some Maneuver code cleanup (commit: 6514738) (details)
146. Fix autostage limit bug that broke coasts (commit: 705c6d5) (details)
147. Residuals calcs update every tick (commit: b4dd0e5) (details)
148. Make KSPStage mandatory stage stats column (commit: 26f5256) (details)
149. Update BetterController settings (commit: 7119129) (details)
150. RCS Fuel Flow (commit: 09cd44b) (details)
151. Add RCS stats to the stage stats display (commit: 45710d6) (details)
152. Fix an RCS bug (commit: 6c988a4) (details)
153. Zero residuals before doing RCS calcs (commit: dbe7998) (details)
154. Change some RCS "display" logic (commit: 8621b2e) (details)
155. slightly better behavior in residual threshold calcs (commit: 738f22b) (details)
156. Fix more residuals handling (commit: ad5622b) (details)
157. Fix DeltaV Stats and planes in RF (commit: 09066d2) (details)
158. Fix roundoff handling in RCS calcs (commit: 85d48d4) (details)
159. Fix residuals calcs again (commit: 7f3cd29) (details)
160. Fix some PVG optimizer edge conditions (commit: 83323fd) (details)
161. Additional box pinning in the PVG optimizer (commit: 4c3a8f2) (details)
162. Better PVG+StageStats integration (commit: 26acae4) (details)
163. Tweak DecouplingAnalyzer algorithm (commit: cbd3964) (details)
164. Add option to override warp to plane (commit: 62f5364) (details)
165. add a comment (commit: 86bf1b3) (details)
166. The barest start at mass costate integration (commit: 2002ba0) (details)
167. Use System.Math statically (commit: be01ad9) (details)
168. ChangeOrbitalElements now uses Forward AutoDiff via Dual numbers/vectors (commit: 89196a3) (details)
169. ReturnFromMoon improvements (commit: 7992f60) (details)
170. Convert ReturnFromMoon maneuver to use analytical Jacobian (commit: dc1bf95) (details)
171. Cleanup ReturnFromMoon (commit: 7624a87) (details)
172. More ReturnFromMoon cleanup (commit: d78272a) (details)
173. ReturnFromMoon: improve final accuracy + allow larger stepsize (commit: f03a1f6) (details)
174. Better RK initial stepsize (commit: e69fdfd) (details)
175. Wire up Hstart in the IVP API (commit: 4287b09) (details)
176. ReturnFromMoon tweaks (commit: 56d7513) (details)
177. Move Ellipticize and Circularize into MJLib (commit: 127d704) (details)
178. Use ChangeInclination out of MJLib (commit: ac2207c) (details)
179. Fix namespace collisions in the MJLib Tests (commit: f673b88) (details)
180. Move MJLib.Statics up a level for convenience (commit: 4806676) (details)
181. MJLib version of HeadingForLaunchInclination (commit: 0fb3a01) (details)
182. remove now used method (commit: 951d135) (details)
183. Add a README.md file for MJLib (commit: 1e5bbc8) (details)
184. I put the README in the wrong spot (commit: b7c667d) (details)
185. Updates ans spell checks (commit: 7fc6293) (details)
186. forgot a word (commit: 2de4267) (details)
187. Remove old FuelFlowSimulation (commit: f5eda67) (details)
188. Hohmann/Bi-impulsive overhaul (commit: a926641) (details)
189. Fix Hohmann bugs and implement more features (commit: 78f7d3c) (details)
190. Fix planning of insertion burn (commit: 83121cc) (details)
191. Fix the FFS root part (commit: 9335acd) (details)
192. Add a warning to the hohmann operation, add note to decoupling analyzer (commit: c411cd8) (details)
193. add time selector to two-impulse transfer (commit: 82b714e) (details)
194. Lots of code cleanup (commit: 684b65e) (details)
195. Fix thrust limiter (commit: c6aa665) (details)
196. Rename ThrustPercentage to ThrottleLimiter (commit: b6ea7de) (details)
197. Make PVG converge under more extreme circumstances (commit: 9f4d60a) (details)
198. Rename to Two-Impulse Maneuver and support fixed time (commit: ed96b5d) (details)
199. Start of refactoring of node executor (commit: 87f6ef2) (details)
200. Node executor code cleanup (commit: e047769) (details)
201. Node executor:  more tweaks to algorithm (commit: c79001f) (details)
202. Behavioral changes to Node Executor (commit: d7a69e0) (details)
203. Fix stock burn termination and bring back MANEUVER_COT (commit: 64da2d1) (details)
204. Fix thrust curves yet again (commit: 6cc7fd1) (details)
205. Fuse principia + stock maneuver node execution behavior (commit: 04245a9) (details)
206. Remove principia node execution button (commit: c441e6b) (details)
207. maybe fix principia node execution (commit: 0162ec3) (details)
208. Node Executor State Machine + RCS burns (commit: f87468c) (details)
209. Fix multinode execution for stock (commit: 437f551) (details)
210. Fix principia execution (commit: 0d2025b) (details)
211. Warp past 10 minutes if we're already aligned (commit: 866e7be) (details)
212. Fix principia multi-node (commit: e21bb53) (details)
213. Run gravity losses continuously (commit: 5826137) (details)
214. FlightRecorder automated cleanup (commit: 5a71aa1) (details)
215. CachedLocalizer cleanup (commit: e60798f) (details)
216. DisplayModule cleanup (commit: 4bff466) (details)
217. Clean up GLUtils (commit: 4bcd32b) (details)
218. Fix multiple enumeration in InstallChecker (commit: 2f04821) (details)
219. Add SerializeField to private Persistent fields (commit: c504550) (details)
220. Make all Persistent fields Public (commit: a0b239b) (details)
221. Rename DisplayModule fields (commit: 6ea3842) (details)
222. cleanup attitude controllers (commit: 37b09f4) (details)
223. GuiUtils cleanup (commit: 6099021) (details)
224. WaypointWindow cleanup (commit: d04257e) (details)
225. Use string interpolation (commit: 95ae54a) (details)
226. Make fields readonly (commit: faf0377) (details)
227. Use TryGetValue (commit: a060fd7) (details)
228. Use TryAdd (commit: dad84bf) (details)
229. Inline out variable declaration (commit: 088be36) (details)
230. Remove unused local variables in Tests (commit: 824cb30) (details)
231. Remove empty statements (commit: e58d5d2) (details)
232. Revert back to old naming (commit: dc4525e) (details)
233. Mark all info items as implicitly used (commit: 89ead30) (details)
234. Fix Principia Node execution again (commit: 0e54ef9) (details)
235. Small fix to the Updater for null attach part (commit: b37673f) (details)
236. Fix node executor ullage (commit: 3b4e45a) (details)
237. Node executor: aligned is good enough for RCS ullaging (commit: e93ea6e) (details)
238. also only use aligned in deciding to start burn (commit: 0af2ef4) (details)
239. Simplify RCS ullage (commit: f06acb0) (details)
240. Add back a minimum rcs time and ignition delay (commit: de9691b) (details)
241. Use propellantStability directly (commit: 586dc2c) (details)
242. Ullage in the Node Executor for at least 1/2 second (commit: bcb686a) (details)
243. The final RCS ullage fix (commit: 50fae03) (details)
244. Tune RCS ullaging more (commit: cd846e3) (details)
245. Add forced ullaging right before ignition (commit: fa93eac) (details)
246. Adding Izzo's Lambert method (commit: c9a4c4e) (details)
247. MechJebLib reorganization (commit: 4b468e2) (details)
248. Change IsFinite() to static from extension (commit: a3a9ff8) (details)
249. Extract MechJebLib as separate project/dll (commit: 7d329fd) (details)
250. Add a newton's method that fails back to bisection (commit: 7d1b6a8) (details)
251. More cleanup and reorg (commit: a264270) (details)
252. Automated cleanup (commit: 5813571) (details)
253. Use WithInheritors to cleanup a lot of jetbrains spam (commit: cec19be) (details)
254. ReturnFromMoon: don't force the meetpoint to be halfway (commit: 675f7a2) (details)
255. ReturnFromMoon: change periapsis constraint (commit: bf9e8dc) (details)
256. ReturnFromMoon updating tests (commit: 805aecd) (details)
257. Add Dormand-Prince 8(5)3 method (commit: 96a998d) (details)
258. Nuke more of the old gravity turner code (commit: 05475d6) (details)
259. ReturnFromMoon: increased precision and logging (commit: 56031e9) (details)
260. add the Tsit5 method (commit: 423d774) (details)
261. Overhaul of ReturnFromMoon initial guesser (commit: a61ca04) (details)
262. ReturnFromMoon: fix dtmin bug well enough for now (commit: 50f1b11) (details)
263. Rewrite and simplify ReturnFromMoon (commit: 3dee06c) (details)
264. ReturnToMoon: the quite stable solution (commit: 201c3d0) (details)
265. Fix ascent-type menu selector (commit: a50fce6) (details)
266. Try to remap old ascent settings to not throw cast errors (commit: 9616391) (details)
267. Rename the AscentTypeInteger and remove old GT params (commit: 7e72dab) (details)
268. Fix custom window presets (commit: 9806024) (details)
269. Beginning of Directory.Build.props settings (commit: fab2471) (details)
270. Fix the longstanding launch to plane bug (commit: fa685a5) (details)
271. Run optimizer during final countdown (commit: dde57e9) (details)
272. Update README.md (commit: d725997) (details)
273. Convert PVG to SQP solver and upgrade alglib (commit: a75e950) (details)
274. Ascent optimizer algorithm changes (commit: eba4650) (details)
275. Fix some new ascent algorithm bugs (commit: 530ff86) (details)
276. Eliminate mass costate entirely from PVG calcs (commit: b6ee029) (details)
277. More PVG test fixes (commit: 0a5f2b8) (details)
278. Multiple PVG fixes (commit: 8b1fb75) (details)
279. Remove unrestartable dead engines from deltaV display (commit: d66ad99) (details)
280. Move StagingController check for dead engine into MJLibBindings (commit: b99cb00) (details)
281. Change boolean-returning API to look more boolean (commit: 64c77bc) (details)
282. Update VesselState.cs (commit: d1c0220) (details)
283. Fix node executor in stock to not RCS ullage (commit: 07ff528) (details)
284. Don't run RCS in Node Executor when IDLE (commit: 33e8258) (details)
285. Some fixes for stock node execution behavior (commit: 66d4fca) (details)
286. Node executor shouldn't control attitude in warp (commit: 207f4ee) (details)
287. Fix issues with RCS being enabled when doing node execution (commit: 991442f) (details)
288. Switch adv transfer to SQP from AUL (commit: cd58b3f) (details)
289. Absolute minimum stepsize of 10 ULPs in ODE solver (commit: 4b50512) (details)
290. Faster ValueInfoItem (commit: 411bc9f) (details)
291. Fix phase angle from mark calculation (commit: 5aa5d7f) (details)
292. Calculate phase angle to target in target's plane (commit: a96d6f1) (details)
293. Log time spent registering info items (commit: 4d26b0f) (details)
294. More instrumentation and caching of InfoItems (commit: fe3eca8) (details)
295. Fix ToSI for floats (commit: f935319) (details)
296. switch bettercontroller back to 90 degree pid (commit: d2e52ef) (details)
297. Add MultFlow to FuelFlowSimulation (commit: 783a863) (details)
298. PID rewrite (commit: cafcaa4) (details)
299. Add missing Profiler.EndSample() (commit: 029ba55) (details)
300. alglib upgrade to 4.03 (commit: 863a378) (details)
301. use genipm intead of ipm (commit: e9a2d95) (details)
302. Move some PVG logging to DEBUG-only (commit: 2c0df94) (details)
303. Fix PVG settings stage numbers (commit: 23b28b6) (details)
304. Only check for ullage on ModuleEnginesRF (commit: b54372b) (details)
305. typo (commit: 83693bc) (details)
306. RCS fuel flow fixes and missing last stage in the VAB (commit: 30d949a) (details)
307. revert back to previous current stage behavior (commit: 804cca3) (details)
308. FFS: always reset current stage if it is too high (commit: 1014850) (details)
309. Revert "revert back to previous current stage behavior" (commit: 8ceb3f8) (details)
310. Code cleanup in MJLib (commit: 4cd92a2) (details)
311. More code cleanup and editorconfig tweaks (commit: 2570ac4) (details)
312. PVG needs to immediately autostage (commit: 1b37703) (details)
313. Also kill throttle before autostaging in PVG (commit: b37d7b3) (details)
314. Only ever insert one coast (commit: 1eba9de) (details)
315. Fix coasts after the early shutoff stage (commit: 8824457) (details)
316. upgrade alglib to 4.04 (commit: fecb434) (details)
317. Fix a problem where coast time doesn't count down (commit: 5f213f6) (details)
318. Fix thrustForward calc for hotstaging (commit: 0857eac) (details)
319. Fix Suicide Burntime calculation (commit: ff69e93) (details)
320. Average initial and final gravity in suicide burn calc (commit: 157e567) (details)
321. restore pid settings to before cafcaa40c1bd3e01283d2970d70e8480267fb48e (commit: c46dc9d) (details)
322. fix builds (commit: 7e3c445) (details)
323. better gate terminal_staging (commit: d93db20) (details)
324. tweak thrust-fault early termination issue (commit: 86deab4) (details)
325. Revert "restore pid settings to before cafcaa40c1bd3e01283d2970d70e8480267fb48e" (commit: 060ef0a) (details)
326. re-revert the pid tuning values (commit: 28031c3) (details)
327. Fix classic launch inclination (commit: 1d5e71f) (details)
328. continually reset pids on the launch pad (commit: e3e631e) (details)
329. Stock ascent cleanup and fixes (commit: d63a6fe) (details)
330. prevent PID windup during PVG staging/spoolup (commit: 6f6f0f0) (details)
331. Add a little margin to the antiwindup in the PVG controller (commit: 97688b2) (details)
332. properly freeze ascent in stock guidance (commit: fbeb9ec) (details)
333. simplify+fix classic ascent corrective steering (commit: 62b9399) (details)
334. fix staging during terminal properly (commit: 6a0dcb0) (details)
335. Fix PVG roll wiggles on vertical ascent (commit: 07f7f2c) (details)
336. Update BetterController to have a real PID for the position (commit: 1b95157) (details)
337. Make some BetterController output more readable (commit: 827317b) (details)
338. Launch Ascent tweaks (commit: e76b323) (details)
339. Don't constantly reset the BetterController PID (commit: f02077c) (details)
340. code cleanup and a couple comments (commit: cfc5e8d) (details)
341. Convert attitude controllers to QuaternionD (commit: 3de2f78) (details)
342. implement still-missing QuaternionD functions (commit: ecedc86) (details)
343. Reformat and cleanup of attitude controllers (commit: 8856a35) (details)
344. Fix saving and loading of vessel-specific settings (commit: 5e920de) (details)
345. Add Vector3d Slerp/Lerp methods (commit: dd5be24) (details)
346. Add DirectionTracker and wire it up to BetterController (commit: f527fad) (details)
347. Run Qα limiter during pitch program again (commit: 29791bd) (details)
348. remove some logspam (commit: 9373fb9) (details)
349. Ignore un-stageable parts (commit: ac1eac4) (details)
350. Update issue templates (commit: 43f5041) (details)
351. Delete .github/ISSUE_TEMPLATE/BUG_REPORT.md (commit: 2c93915) (details)
352. Delete .github/ISSUE_TEMPLATE/HELP_REQUEST.md (commit: 21c2973) (details)
353. Delete .github/ISSUE_TEMPLATE/LOCALIZATION_SUPPORT.md (commit: 860cef8) (details)
354. Delete .github/ISSUE_TEMPLATE/NEW_KSP_VERSION_SUPPORT.md (commit: cf60419) (details)
355. Some more stage numbering fixes (commit: 79fd6ea) (details)
356. Update bettercontroller tuning (commit: 7533747) (details)
357. Get the gimbal rate into the vessel state (commit: 6bfabbb) (details)
358. better controller better ui (commit: badd66c) (details)
359. Minor BetterController tweaks (commit: 999de9e) (details)
360. Prevent DecelerationBurn from setting the throttle to 0 after PDI (commit: 6aef833) (details)
361. Change braking burn end altitude to 200m (commit: a30d52a) (details)
362. Deceleration burn warp fixes (commit: 94dbb19) (details)
363. Only warp when angular velocity is near zero (commit: 8f6728b) (details)
364. Attempt to prevent full engine shutdown during vertical final descent. (commit: ea9a563) (details)
365. Make min throttle limiter follow global min throttle limiter (commit: 8d974b4) (details)
366. Fix for attitude excursion in vertical descent (commit: d23e55e) (details)
367. Add a fudge factor for control surface response (commit: 424d5ed) (details)
368. Better Test for V3 APIs (commit: f8b777d) (details)
369. Add V3 tests for the component analysis APIs (commit: 439096b) (details)
370. Add RequestActiveThrottle() API to ThrustController (commit: d39820d) (details)
371. Attempt to implement new RequestActiveThrottle API. (commit: e227003) (details)
372. Small untested fix attempt (commit: 834f8c5) (details)
373. Add more V3 API tests (commit: b52c752) (details)
374. That's what I get for not even test compiling. Optional arguments don't work that way in C# (commit: b6e7238) (details)
375. Actually fix engine shutoff (commit: 32265bd) (details)
376. Add tests for V3 sph2cart cart2sph and xzy (commit: 50f9af5) (details)
377. add some invalid value checking around SafeAsin/Acos and tests (commit: 891713d) (details)
378. Add more tests around V3 class (commit: e63426b) (details)
379. Changes to final descent guidance and throttle logic to fix attitude excursions. (commit: 6a986a7) (details)
380. Update alglib to 4.06 (commit: f2ae914) (details)
381. Fix hotstaging without Δv window open (commit: d40ca8a) (details)
382. Change the logic for speed control when in final descent below 300m with more than 5 m/s horizontal velocity. (commit: 571fe4f) (details)
383. Remove all throttling logic and just burn at full until the horizontal speed is sufficiently reduced. (commit: ed19fe7) (details)
384. Update version to 2.15.0 (commit: 7e9ed8a) (details)

WIP PVG rewrite

Signed-off-by: Lamont Granquist <lamont@scriptkiddie.org>
 (commit: dc6464b)

more massive refactoring

Signed-off-by: Lamont Granquist <lamont@scriptkiddie.org>
 (commit: 1b79e0c)

checkpoint

Signed-off-by: Lamont Granquist <lamont@scriptkiddie.org>
 (commit: 08b233a)

push an update

Signed-off-by: Lamont Granquist <lamont@scriptkiddie.org>
 (commit: 329d8fe)

another checkpoint

optimization is now done in a background thread

Signed-off-by: Lamont Granquist <lamont@scriptkiddie.org>
 (commit: 8a375f9)

checkpoint again.

whatever i got done before vacation.

Signed-off-by: Lamont Granquist <lamont@scriptkiddie.org>
 (commit: a37876e)

WIP

Signed-off-by: Lamont Granquist <lamont@scriptkiddie.org>
 (commit: c728bac)

mostly working

Signed-off-by: Lamont Granquist <lamont@scriptkiddie.org>
 (commit: d4d90db)

more WIP

Signed-off-by: Lamont Granquist <lamont@scriptkiddie.org>
 (commit: 342d5a0)

many fixes

the analytic integrator now supports unguided stages properly i think

Signed-off-by: Lamont Granquist <lamont@scriptkiddie.org>
 (commit: 0e68498)

Code cleanup

Signed-off-by: Lamont Granquist <lamont@scriptkiddie.org>
 (commit: 51c6c15)

fixed burn time fixes and test fixes

Signed-off-by: Lamont Granquist <lamont@scriptkiddie.org>
 (commit: 3dfe8a1)

Fix optimizer trying to coast when it shouldn't

Signed-off-by: Lamont Granquist <lamont@scriptkiddie.org>
 (commit: 46d5fd7)

Some code cleanup in the PVG Ascent algorithm

breaks up the algorithm into fixed-vs-optimized rocket burntimes since
those are getting difficult to handle, and also starts to treat the
phase list a bit more immutably since that was getting crazy.

Signed-off-by: Lamont Granquist <lamont@scriptkiddie.org>
 (commit: 3c4b656)

unrelated code cleanup

Signed-off-by: Lamont Granquist <lamont@scriptkiddie.org>
 (commit: 57b0624)

Fix FuelFlowSim for NO_FLOW resources in Engines

This fixes a bug affecting Delta IV Heavy setup with asparagus staging.  The RS68 engines have had Ablator added to them as an additional "fuel" which is NO_FLOW.

The old check CanDrawResource() which was used to determine if staging was allowed was not aware of fuel modes and that caused it to check if any part in the crossfeedset had the fuel even for NO_FLOW fuels.  This meant that after the side boosters flamed out that the center core would still be burning Ablator and it would that Ablator "accessible" in its crossfeed set coming from the side engines (because of asparagus).

This may also affect SRBs since those fuels are NO_FLOW as well?

However, I think it requires something like Asparagus staging And given that RO doesn't allow crossfeed through decouplers or fuel pipes it may be a pretty narrow bugfix.

Signed-off-by: Lamont Granquist <lamont@scriptkiddie.org>
 (commit: 46a9693)

Fix StagingController for NO_FLOW fuels

Companion to #1612 to fix autostaging.

Had to sort of invert the algorithm to loop over all the propellants in
all the active engines so that I could grab the Mode off of the engine
propellant (seems not to be associated with the resources on the tank).

This worked on my test Delta IV Heavy on the first try...  Makes me
wonder what I broke...

Signed-off-by: Lamont Granquist <lamont@scriptkiddie.org>
 (commit: 42435f4)

remove dead code comment

Signed-off-by: Lamont Granquist <lamont@scriptkiddie.org>
 (commit: 93f2709)

simplify MJlib logger

i was doing way too much work here.

Signed-off-by: Lamont Granquist <lamont@scriptkiddie.org>
 (commit: 9a0d17f)

Use new ToSI extnesion method everywhere

Should be backwards compatible only with bugfixes and tests.

Signed-off-by: Lamont Granquist <lamont@scriptkiddie.org>
 (commit: ba84b68)

Try to see if JetBrains.Annotations breaks builds

Signed-off-by: Lamont Granquist <lamont@scriptkiddie.org>
 (commit: 97b02b9)

Apply UsedImplicitly to fix dead code analysis

MJ uses a bunch of reflection.

Signed-off-by: Lamont Granquist <lamont@scriptkiddie.org>
 (commit: e62cd5e)

code cleanup to MechJebCore

Signed-off-by: Lamont Granquist <lamont@scriptkiddie.org>
 (commit: be96049)

Use aggressive inlining

This makes a large difference, around a factor of 3x speedup in
stuff like Deg2Rad() on my machine.  There is still a tiny perf
hit of around 5% compared to just inlining the math directly,
which seems perfectly acceptable.

I didn't exhaustively perf check all these APIs, but they're all
trivial enough that the codebloat should be small.  Stuff like
ToSI() would clearly cause bloat and so are not included.

Signed-off-by: Lamont Granquist <lamont@scriptkiddie.org>
 (commit: df5aebc)

Code cleanup for the Maneuver Operations

This might be a breaking change if anyone out there uses these
via reflection or whatnot.

OperationAdvancedTransfer still needs more love.

Signed-off-by: Lamont Granquist <lamont@scriptkiddie.org>
 (commit: 12ae2c6)

Move ODE in MathJ to MechJebLib

Signed-off-by: Lamont Granquist <lamont@scriptkiddie.org>
 (commit: 2552971)

Change License on ODE Solver

This was all my work originally.

Signed-off-by: Lamont Granquist <lamont@scriptkiddie.org>
 (commit: 447f766)

Move "SpaceMath" function into MechJebLib

- the singleimpulsehyperbolicburn function still allocates due to a closure.
- functions probably need some better organization at this point.

Signed-off-by: Lamont Granquist <lamont@scriptkiddie.org>
 (commit: f5a5d40)

M3 fixes and a lot more

- M3 test coverage
- M3 fixes (matrix inverse, transpose and other fixes)
- M3 documentation
- move PIDLoop to MechJebLib
- tweak NearlyEqual to make more sense near zero
- bit more test coverage for some other Functions and APIs

Signed-off-by: Lamont Granquist <lamont@scriptkiddie.org>
 (commit: ff93427)

Change "attach altitude" to "burnout altitude"
 (commit: 47daf5b)

Change "stop at stage" to make it clearer it affects autostaging
 (commit: ad2c710)

remove spammy logging
 (commit: eb80ae9)

Address warnings

This test I find intellectually intersting as living documentation
but causes too many warnings.
 (commit: d28bbe9)

Fix PVG terminal staging bug

If the insertion stage has RCS but no relights then the
guidance controller will wind up staging off the payload
once terminal RCS happens (breaking the terminal RCS).

This reverts to previous behavior that I didn't realize
was important.

Signed-off-by: Lamont Granquist <lamont@scriptkiddie.org>
 (commit: adf3ecd)

Fix the autostage limit

Somehow I completely deleted the functionality.

Signed-off-by: Lamont Granquist <lamont@scriptkiddie.org>
 (commit: 574c150)

Track down and fix some annoying allocations in the ODE+interpolants

The ODE should be allocation-free now via memory pooling.

Signed-off-by: Lamont Granquist <lamont@scriptkiddie.org>
 (commit: 9763ffd)

Add proper tests for Gooding Lambert calculator

At least have pretty good coverage of single-revolution use case now

Signed-off-by: Lamont Granquist <lamont@scriptkiddie.org>
 (commit: 2be3676)

Remove SwapYZ API

Just use KSP's Vector3d.xyz field directly.

Signed-off-by: Lamont Granquist <lamont@scriptkiddie.org>
 (commit: 81232eb)

Automated cleanup of many ancillary source code modules

This all probably needs a second pass by a human to make it
all nicer, this is only what the robot thinks is good.

Signed-off-by: Lamont Granquist <lamont@scriptkiddie.org>
 (commit: 78d2fa5)

Rename the "Swapped" API

I never remember if "Swapped" in the API refers to world
space being "Swapped" or right-handed space being "Swapped".

Removes Prograde() and Up() since those are now actual KSP
APIs and those extension methods are confusingly overwriting
what KSP does (although both are world coordinates and
backwards compatible -- I hope).

Signed-off-by: Lamont Granquist <lamont@scriptkiddie.org>
 (commit: 6086f8e)

clean up the DeltaVToInterceptAtTime() API.

adds some glue APIs to help dealing with converting between
KSP APIs and the right handed V3 APIs.

smears some aggressive inlining across many tiny APIs which
should be beneficial, but i haven't tested.

Signed-off-by: Lamont Granquist <lamont@scriptkiddie.org>
 (commit: bb548ca)

fix probable transfer calculator bug

This has to be buggy, and its easier to see now that I renamed
the APIs to included "Rotated" in them.  We don't rotate to start
so we shouldn't be rotating to end.

Signed-off-by: Lamont Granquist <lamont@scriptkiddie.org>
 (commit: 2061c32)

ReturnFromMoon reimplementation and large reorganization

- Reorganizes a ton of code, I'm still only halfway done and still
  not at all happy with it.

- DeltaVToChangePeriapsis/Apoapsis have been reimplemented with a
  MJLib version that solves an SQP problem to provide an optimal
  solution.  I still need to play around with making it 2D by
  converting to the perifocal frame and using analytic derivatives.

- DeltaVAndTimeForMoonReturnEjection got a massive reimplementation
  using Sims-Flanagan transcription to turn it into an SQP problem.
  The result should be nearly optimal and much more accurate.  The
  remaining inaccuracies are dominated by my simulation and KSP not
  agreeing on something.  Probably something like the exact way
  that KSP executes an SOI transition.

Somewhere in here I think I fixed things like the MJLib
TimeToNextRadius() implementation for hyperbolic orbits.

Signed-off-by: Lamont Granquist <lamont@scriptkiddie.org>
 (commit: 4bb4091)